Practical conversion of artemisinic acid in desoxyartemisinin - The

Tara Sprules, Michael R. Carpenter, Paul D. Cotter, Colin Hill, R. Paul Ross, and ... Si-Hyung Park, Cindy K. Angerhofer, Geoffrey A. Cordell, and...
0 downloads 0 Views 431KB Size
J. Org. Chem. complement to previously reported methods. Bichalcones serve t u important synthetic intermediates in the syntheses of various biflavonoids differing in the oxidation level of the C3 unit. Our procedure should therefore be applicable to nearly all natural biflavonoids through the intermediacy of bichalcones prepared from suitably substituted chalcones. Moreover, our results indicate that iodine in alkaline methanol is a simple and very useful reagent for phenol oxidative coupling. Its synthetic utility could well be exploited in those instances where the phenolic groups are hydrogen bonded. The hydrogen-bonded hydroxyls have been reported7 to resist oxidation.

1986,51, 5417-5419

5417

contents were stirred at room temperature for 3 h and poured into ice-cold water. The organic material was extracted with ethyl acetate and dried over anhydrous sodium sulfate. Evaporation of the solvent gave a yellow solid, which was crystallized from alcohol to give crystals of 4 (370 mg, 74%), mp 180-181 “C; UV A- (MeOH) 210,230,330,380 nm; IR u, (KBr) 3440,1620 cm-’; NMR (CDC1,) 6 3.85, 3.90, 4.00, 4.03 (3 H, each, s, 4,4‘,4“,4”‘OCH3), 6.65 (1 H, s, 3”’-H), 6.90 (7 H, mc, 5’,(~,(~’,3,5,3”,5’’-H), 7.75 (7 H, mc, 6’,P,p’,2,6,2’’,S”-H), 8.30 (1 H, s, 6’”-H), 13.60 (1 H, s, 2’-OH), 14.00 (1 H, s, 2”’-OH); MS, m/z (relative intensity) 567 (3.21), 566 (M, 19.23), 565 (30.64),459 (4.62),458 (3.29), 433 (14.19), 432 (19,61), 405 (45.23), 404 (67.32), 326 (16.73), 325 (28.11), 299 (18.31), 271 (8.67), 163 (7.35), 161 (67.32), 150 (1 4.59), 149 (10.95), 135 (24.45), 134 (loo), 133 (34.56), 121 (28.91).

Experimental Section Acknowledgment. S.M.A. acknowledges financial asMelting points were determined on a Kofler microscopical hot sistance from the Council of Scientific and Industrial stage and are uncorrected. IR and UV spectra were obtained on Research, New Delhi. Pye Unicam SP3-100 and PU 8800 spectrophotometers, respecRegistry No. 1, 3420-72-2; 2, 57290-97-8; 3, 2198-19-8; 4, tively. NMR spectra were measured with a Varian A-60D in105281-48-9; phloroacetophenone 4,6-dimethyl ether, 90-24-4; strument, and mass spectra with a JEOL JMS-D300 instrument. 2-hydroxy-4-methoxyacetophenone, 552-41-0; p-anisaldehyde, Preparation of 2’-Hydroxy-4,4’,6’-trimethoxychalcone ( 1). 123-11-5; succedaneflavone hexamethyl ether, 57290-98-9. To a mixture of phloroacetophenone-4,6-dimethylether (5 g, 25.5 mmol) and p-anisaldehyde (3.47 g, 25.5 mmol) in alcohol was added a hot 50% aqueous solution of sodium hydroxide (10 g). The resulting mixture was heated at 50 “C for 30 min. The contents were then cooled and poured into cold water and neuPractical Conversion of Artemisinic Acid into tralized with dilute HCl. A yellow solid was obtained that was Desoxyartemisinin filtered, washed, and dried. Crystallization from alcohol yielded yellow needles of 1 (5.85 g, 73%), mp 112-114 “C (lit.14mp 113-114 Mankil Jung,*’ Hala N. ElSohly, and Edward M. Croom2 “C). Research Institute of Pharmaceutical Sciences, School of Preparation of 2’-Hydroxy-4,4’-dimethoxychalcone(3). To Pharmacy, University of Mississippi, a cold suspension of equimolar amounts of 2-hydroxy-4-methUniversity, Mississippi 38677 oxyacetophenone (8 g, 50 mmol) and p-anisaldehyde (7 g, 50 “01) in alcohol (160 mL) was added a cold 60% aqueous solution Andrew T. McPhail and Donald R. McPhail of potassium hydroxide (95 g). The flask was stoppered securely Department of Chemistry, Paul M.Gross Chemical and allowed to stand at room temperature for 1 week with ocLaboratory, Duke University, casional shaking. The contents were then poured on to crushed Durham, North Carolina 27706 ice and neutralized with dilute HC1 to give a yellow precipitate. The crude solid was filtered, washed with water, dried, and Received June 3, 1986 crystallized from methanol to afford yellow needles of 3 (10.4 g, 74%), mp 89-91 “C (lit.7 mp 90-91 “C). Artemisinin (Qinghaosu, 2) isolated from Artemisia 2’,2”’-D~~ydroxy~~,4’,4’’,4”’,6’,6’’’-hexamethoxy[5’,5’’’]biannua has recently been used in China as a new type of chalcone (2). 2’-Hydroxy-4,4’,6’-trimethoxychalcone (314 mg, antimalarial drug with rapid action and low toxicity against 1 mmol) was dissolved in methanol, and a methanolic solution chloroquine-resistant malaria.3i4 Artemisinin is a novel of potassium hydroxide (0.5 g) was added to it. To the resulting sesquiterpene lactone bearing an unusual cyclic peroxide mixture was added iodine (127 mg, 0.5 mmol) with shaking. The function. The combination of an interesting biological contents were stirred at room temperature for 3 h and poured into ice-cold water. A red solid precipitated that was filtered, activity, a novel chemical structure, and a low yield from washed with water, and dried. Crystallization from alcohol yielded natural sources prompted us to search for a new synthesis red crystals of 2 (220 mg, 70%), mp 189-191 “C (lit.13mp 191-192 of artemisinin and related compounds. Although inter“C): W A, (MeOH) 228,350,370 nm; IR,,v (Kl3r) 3440,1625, esting ~ y n t h e s e s ~of- ~artemisinin and desoxyartemisinin 1600 cm-’; NMR (Me2SO-d,) 6 3.80-4.00 (18 H, s, were reported recently, these complex syntheses do not 4,4’,4’’,4’’’,6’,6’”-OCHJ,5.62 (2 H, S, 3’,3”’-H), 6.90 (4 H, d, J = provide feasible methods for large-scale production. A. 9 Hz, 3,5,3”,5”-H), 7.50 (4 H, d, J = 9 Hz, 2,6,2”,6”-H), 7.65 (4 annua has been found to contain approximately 8-10 times H, s, a,P,a’,p’-H), 14.35 (2 H, s, 2’,2”’-OH); MS, m/z (relative more artemisinic acid than artemisinin.8 We therefore intensity) 626 (M, 80.68), 611 (8.62), 519 (3.56), 493 (6.54), 492 attempted a new synthesis that embodies highly stereo(13.56), 491 (lO.lO), 465 (15.48), 464 (22.49), 463 (56.86), 461 (20.19),358 (16.42), 357 (3.14), 327 (21.43),314 (20.24), 313 (8.92), specific conversion of artemisinic acid (1)into artemisinin 312 (11.73), 301 (11.64), 194 (16.37), 179 (12.16), 180 (4.50), 161 (2) (Scheme I). Stereoselective reduction7 of artemisinic (19.41), 134 (60.21), 133 (22.85), 121 (100). acid (1) [LiBH4 (5.3 equiv), NiC12.6H20 (0.49 equiv), Oxidation of 2. A mixture of 2 (100 mg) and Se02 (0.5 g) in CH30H, room temperature, 2 h] into the dihydro comisoamyl alcohol was refluxed for 24 h. The contents were then pound 3a (quantitative yield), followed by ozonolysis7 of poured on the crushed ice. A solid precipitated that was filtered, 3a (03, CH2C12/CH30H= 1/1.5, -78 “C for 1 h and then washed with water, and dried. Crystallization from a chloroCH3SCH3 workup), gave the keto aldehyde 4, 75%. form/methanol mixture gave succedaneailavone hexamethyl ether. Melting point, TLC, and NMR spectrum were in good agreement with that reported in the 1iterat~re.l~ (1) Department of Medicinal Chemistry, University of Mississippi. 2‘,2””Dihydroxy-4,4‘,4‘‘,4‘‘‘-tetramethoxy[ 3‘,5“‘]bichalcone (2) Department of Pharmacognosy, University of Mississippi. (3) Qinghaosu Antimalaria Coordinating Research Group Yaoxue (4). The reaction was performed in the previously described Tongbao 1979,14,49. manner. Thus, 2’-hydroxy-4,4’-dimethoxychalcone(500 mg, 1.76 (4) Qinghaosu Antimalaria Coordinating Research Group Chin. Med. mmol) was dissolved in methanol and a methanolic solution of J . 1979, 92. potassium hydroxide (1 g) was added to it with shaking. Iodine ( 5 ) Schmid, G.; Hofheinz, W. J.Am. Chem. SOC.1983, 105, 624. (112 mg, 0.88 mmol) was added to the resulting mixture and the (6) Xu, X.; Zhu, J.; Huang, D.; Zhou, W. Acta Chim. Sin. 1983,41,574. (14) Geissman,

T. A.; Clinton, R. 0. J . Org. Chem. 1946, 81, 697.

0022-3263/86/1951-5417$01.50/0

(7) Xu, X.; Zhu, J.; Huang, D.; Zhou, W. Tetrahedron 1986, 42,818. (8)ElSohly, H. N.; Croom, E. M.; ElSohly, M. A.; Jung, M., unpublished result.

0 1986 American Chemical Society

5418 J. Org. Chem., Vol. 51, No. 26, 1986

Notes

Scheme I"

COpH 1

3a, R = H b . R=CH3

4

0

5

6

2

"Key: (a) LiBH, (5.3 equiv), NiC1,-2Hz0 (0.49 equiv), CH,OH, room temperature; (b) 0,, CH,Cl,/CH,OH = 1/1,5,-78 O C , 1 h, then CHaSCH, workup; (c) 70% HC104, T H F / H , O = l j l , room temperature, 7 h.

O(17)

Figure 1. ORTEP plot of the structure of 11, small circles denoting hydrogen atoms.

structural analogues of artemisinin. Thus, epoxidation (m-CPBA, CHCl,, 0 "C, 1 h) of 7 gave a sesquiterpene lactone 11 in one step79l3(64% 1. We rationalize this result as follows: preliminary stereospecific a-face addition of 1 3b oxygen into the double bond of 7 could give the epoxide 8. Epoxide ring opening of 8 to give the 1,2-dioxetane 9 could be followed by cyclic peroxide ring cleavage of 9 to afford the keto aldehyde 10. Finally, facile cyclization of 7 8 10 would result into desoxyartemisinin (1 1). The absolute configuration of 11 a t C-1, C-3, C-5a, C-6, C-8a, C-9, and C-12 was established unequivocally by single-crystal X-ray analysis as l S , 3R, 5aS, 6R, 8aS, 9R, and 12R; thus, three new chiral centers (C-1, C-3, and C-12) were generated in natural configuration. The crystal structure was solved by direct methods.'* Full-matrix least-squares adjustment Q of non-hydrogen atom positional and thermal parameters, 11 with hydrogen atoms included at their calculated positions, converged to R = 0.043 (R, = 0.063).15 A view of the "Key: (a)-(c) see Scheme I; (d) CH2N2, Et20, room temperasolid-state conformation is provided in Figure 1. The ture, 20 min; (e) IOz, methylene blue, CH,OH, room temperature, 1,3-dioxalane ring is in an envelope form with 0-16 as the 2.5 h; (f) m-CPBA, CHCl,, 0 "C, 1 h. out-of-plane atom.16 Endocyclic torsion angles in the Compound 4 bearing a free acid group was successfully &lactone ring16 are related by an approximate mirror plane cyclized (70% HC104,T H F / H 2 0 = 1/1,room temperature, of symmetry passing through C-8a and 0-11, and, with very 7 h) into the enol lactone 5 , 52%. For p h o t o o ~ i d a t i o n ~ , ~ small values for two adjacent angles, this ring has a halfof 5 into artemisinin (2), all attempts of chiral addition of boat (envelope) form. activated oxygen (high mercury arc lamp with methylene Although desoxyartemisinin shows no antimalarial acblue or hematoporphyrin as a sensitizer) via the transition tivity, this is an important metabolite for toxicology state 6 have been unsuccessful. This might be due to studies. This synthesis requires only four steps from decreased electron density of the double bond attached to readily available, optically active starting material 1 and the electron-withdrawing ester group within the enol lacproceeds in 42% overall yield. In conclusion, this process tone ring. During these synthetic studies on artemisinin, demonstrates a new and practical synthesis of desoxyarwe found an interesting new conversion of artemisinic acid temisinin from artemisinic acid. Conversion of artemisinic (1)into desoxyartemisinin (1lyJ3in natural configuration acid and desoxyartemisinin into artemisinin is still under (Scheme 11). Thus, esterification' of artemisinic acid (1) investigation. (CH2N2,ethyl ether, room temperature, 20 min) (98%), followed by stereoselective reduction of a,p-unsaturated (13) Hydrogenation of artemisinin in the presence of a Pd/CaCOs bond, afforded the dihydro artemisinic methyl ester 3b, catalyst causes loss of one of the biologically important peroxide oxygens 95 % . Stereoselective and regiospecific p h o t o o x i d a t i ~ n ~ ~ ~to~ give desoxyartemisinin. See:China Coorperative Research Group on Qinghaosu and Ita Derivatives as Antimalarials J. D a d . Chin. Med. 1982, of 3b (lo2,methylene blue, CH30H, room temperature, 2.5 2(1), 9. Desoxyartemisinin isolated from A. annua is also the biotransh) provided the allylic hydroperoxide11J2 7 as a stable form: formation product in the body. See: Zhu, D.; Huang, B.; Chen, Z.; Yin, oil; 70%. Compound 7 is a versatile intermediate to obtain M. Yao Hsueh Hsueh Pa0 1980,15(8), 509.

*

(9) Marshall, J. A,; Hochstetler, A. R. J. Org. Chem. 1966, 31, 1020. (10) Schultz, A. G.; Schlessinger, R. H. Tetrahedron Lett. 1970, 2731. (11) Xu, X.; Huang, D.; Zhu, J.; Zhou, W. Acta Chim. Sin. 1982,40, 1081. (12) Xu, X.; Zhu, J.; Zhou, W. A c t a Chim. Sin. 1985, 43, 48.

(14) Main, P. MULTANllIBg; University of York: York, England, 1982. (15) R = .EIIFol- IFell/.EI FoI;R , = l.E(l~ol - lFc12)/.E~1~o121"2~ (16) Endocyclic torsion angles (q,) about the bonds between atoms C-i and C - j follow (deg): w1,2 3.8; ~ 2 22.6; , ~~ 3 . -41.5; ~ 6 ( L ' ~ ~ ,42.5; ~ G ~ 1 . 1 2-28.1 in the 1,3-dioxalane ring wl,il 0.5; 01,12 -25.9; w ~ 52.9; ~ , -56.7; ~ ~ w9.10 34.5; w~,,~, -5.3 in the &lactone ring.

J. Org. Chem. 1986,51, 5419-5421

5419

Scheme I

Experimental Section Enol Lactone 5. HC104 (7 mL 70%)was added to the solution of the keto aldehyde 4 (533 mg, 2.0 mmol) in THF (10 mL) and water (10 mL). The whole solution was stirred for 7 h at room temperature. The mixture was partitioned between CHzClz(50 mL) and water (15 mL) and then extracted, washed with brine, and dried (Mg SO4). The evaporation in vacuo gave a colorless oil, which was further purified on column chromatography (silica gel, 8/1= CHC13/CH30H)to afford enol lactone 5 260 mg (52%); 'H NMR (CDCl,, Me4Si)6 6.72 (br s, 1H, enol H), 3.19 (m, 1 H, OZCCH), 2.4 (9, 3 H, CHSCO); IR (CHC1,) 1705 (C=O), 1650 (C=O, enol lactone) cm-'. Allylic Hydroperoxide 7. A stream of pure oxygen was admitted through a gas dispersion tube to a solution of (9R)methyl artemisinate (3b;500 mg, 2.0 mmol) and methylene blue (50 mg, 0.13 mmol) in anhydrous CH,OH (30 mL). The mixture was irradiated by high mercury arc lamp during 2.5 h at room temperature. Evaporation in vacuo and purification by column chromatography (silica gel, 1 / 2 = EtOAc/hexane) afforded the allylic hydroperoxide 7 [401 mg (70%)]as a slightly yellow oil: 'H NMR (CDCI,, Me4Si) 6 5.27 ( 8 , 1 H, CH=), 3.80 (s, 3 H, COzCH,), 2.81 (m, 1 H, CHC02),1.22 (9, 3 H, CH, at C-3); IR (neat) 3400 (OOH), 2940, 2560, 1740 (C=O), 1650, 1440, 1200, 1170, 760 (C=C) cm-'. Desoxyartemisinin (11). To the solution of the allylic hydroperoxide 7 (400 mg, 1.42 mmol) in dry CHC13(6 mL) was added m-CPBA (360 mg, 2.08 mmol). The mixture was stirred at 0 "C for 1 h. Evaporation in vacuo and purification on column chromatography (silica gel, 1/2 = EtOAc/hexane) gave a solid. Recrystallization from cyclohexane afforded desoxyartemisinin (11) [240 mg (64%)]as a colorless crystal: mp 108-111 "C (cyclohexane) (lit.7mp 110-111 "C); [a]25D -99.75" ( c 0.4, CHCl,); 'H NMR (CDCI,, Me4%)6 5.70 (8, 1 H, C-l), 3.21 (m, 1 H, C-Q), 1.53 (s, 3 H, C-3); IR (CHC13) 1740 cm-* (&lactone C=O); MS (70ev) m / e 266 (M'). Anal. Calcd for C15H2204: C, 67.67; H, 8.27; 0, 24.06. Found; C, 67.77; H, 8.33; 0, 23.78.

1

A

Supplementary Material Available: Tables of crystal data, atomic positional and thermal parameters, bond lengths and angles, and torsion angles for desoxyartemisinin (11) (6 pages). Ordering information is given on any current masthead page.

A Proposed Kinetic Method for Improving the Optical Purity of Synthetic Peptides Eugene M. Holleran* and Joseph Kovacs Chemistry Department, S t . John's University, Jamaica, New York 11439 Received April 10, 1986

An important consideration in the synthesis of peptides is the optical purity of the product. Generally, the coupling of smaller peptides or amino acids, themselves optically pure to begin with, produces a product that is not optically pure. This is because the free amine group on one reactant catalyzes the racemization of the other reactant during the coupling. Thus some of the product is in the "wrong" isomeric form. This is a highly undesirable result, and much research has been devoted to minimizing its ex-

E

L

2

G

E

D

~

A

In this paper we explore the improvement of optical purity that might be gained by the simple means of using an excess concentration of the reagent that racemizes. We represent the overall coupling reaction as A E P, where P is the peptide product. E can be an active ester, and A has a free amine group. The ester is originally all in the L form, EL, but during the reaction it is partially converted to the D form, ED,before coupling. Thus some of the product is the undesired diastereomer, PD. We assume the reaction in Scheme I as our working hypothesis. The isomeric forms of P are assumed to be stable and not to interconvert. This scheme differs from both the independent and the dependent parallel reaction schemes discussed in an overall treatment by Ugi, et al.4 We now inquire into the effect of three parameters, r , x , and m, on the purity of the product according to this scheme. The first, r , is the ratio A o / E oof the initial concentrations of amine and ester. T h e second, x , is the fraction of Eo that has reacted by any time, t . (Note that only if r = 1 will x also equal the fraction of A,, reacted.) From these definitions we can express the concentrations of product and reactants as shown in eq 1. (Here, and

+

-

E = Eo(1 - X )

A = Eo(r - x )

Acknowledgment. This research was supported by the Research Institute of Pharmaceutical Sciences, School of Pharmacy, University of Mississippi. We express our thanks to Drs. Farouk El-Feraly, Mahmoud A. ElSohly, and Robert Sindelar for helpful discussions and Mr. Tim West for technical assistance. Registry No. 1,80286-58-4;1 (methyl ester), 82869-24-7;3b, 87391-99-9; 4, 105250-95-1; 5 , 105231-07-0; 7, 85031-63-6; 11, 72826-63-2.

~

(1)

P=E~x subsequently, capital letters may be taken to represent concentrations.) At any time, E = ED EL, and P = PD PL. If the amine is in excess, then r > 1, and upon completion of the reaction, x = 1, E = 0, A = E,(r - l), and P = E,. On the other hand, if the ester is in excess, then r < 1, and a t completion, x = r , E = E,(1 - r ) , A = 0 , and P = rEo = Ao. In this case, the final, maximum value of x is less than one. The third parameter, m, is defined as the ratio, k,/k,, of the racemization rate constant ( k , = 2k1 = 2kJ to the coupling rate constant ( k , = k3 = k4). (We are assuming for simplicity that the two optical isomers of E do not differ in their racemization rates or in their coupling rates.) In general, it is obvious that a small value of m favors product purity. We define the extent of the product's optical impurity, F , as the fraction of the product in the D form, or F = P D / P . We now determine exactly how this product impurity, F , depends on the rate constant ratio, m, the initial concentration ratio, r , and the extent of the reaction as defined by x . T o do this we must first investigate the variation of ED during the reaction, because it is from the D form of E that the D form of P is produced in step 4. T h e rates, R, of the four steps in the reaction scheme can be expressed as shown in eq 2. Later, when we use ratios of these rates, the factor kJ will always cancel out.

+

+

(1)Kemp, D.S. The Peptides; Gross, E., Meienhofer, J., Eds.; Academic: New York, 1979; Vol. 1, p 317. (2) Kovacs, J., in ref 1, 1980; Vol. 2, p 529. (3)Benoiton, N. L.,in ref 1, 1983;Vol. 5,p 218. (4)Brandt, J.; Jochum, C.; Ugi, I.; Jochum, P. Tetrahedron 1977,33, 1353.

0022-3263/86/1951-5419$01.50/0 0 1986 American Chemical Society